On: Reliable novelty: New should not trump true

[knolan]

Whenever I look at a biology paper in Nature I frequently see figures like this:

https://www.nature.com/articles/s41586-019-1112-8/figures/5

They’re utterly incomprehensible to my lowly engineers eye and seem deliberately designed to look as complex as possible.

What do these kinds of images and plots tell us of any value?

[cowsandmilk → knolan]

As someone who did computational biophysics, my opinion is that MD is meaningless and those plots don't tell you anything of value. Most likely, the senior author is friends with a MD expert and wanted an excuse to put the friend on the paper.

[knolan → cowsandmilk]

This type of visualisation appears pretty common in such papers so I think perhaps the authors believe that they now need to include them to get published, a cargo cult of sorts.

[nonbel → knolan]

> "What do these kinds of images and plots tell us of any value?"

It tells you "p38γ is essential for cell cycle progression and liver tumorigenesis". You have to be blind not to see that.